This paper investigates the employment effects of a mass roll-out of household electrification in rural
South Africa. For the 470,000 households that were connected in this area, the new infrastructure
represented an improvement in home production technology. I collected administrative and
geographic data on this expansion and matched it to two waves of aggregate Census data to test for
effects on employment and on fuel use at home. I exploit variation in electricity project placement
and timing to estimate district fixed-effects models, and instrument for project placement using land
gradient that directly affects the cost of grid expansion. My findings show that cooking with wood
falls sharply in treated areas over a five-year period, and lighting and cooking with electricity
increase substantially. IV employment results indicate asymmetric responses by gender: female
employment rates increase by 13.5 percentage points in treated areas, while there are no significant
male effects. Middle-poor communities respond most to the new option to use electricity, and
employment effects are large for women in their thirties and forties who are less constrained by
child-care responsibilities. This new evidence on how home production infrastructure can affect the
extensive margin of work for women contributes to a growing literature on the effects of public
infrastructure in developing countries. The results suggest that studies that do not examine
employment outcomes may miss important economic effects of these infrastructure investments.